7/31/2019 Wc 500027737

1/45

1/45 EMEA 2004

SCIENTIFIC DISCUSSION

This module reflects the initial scientific discussion and scientific discussion on procedures whichhave been finalised before 1 November 2004. For scientific information on procedures after this date

please refer to module 8B.

Introduction

Arixtra is indicated for the prevention of Venous Thromboembolic Events (VTE) in patientsundergoing major orthopaedic surgery of the lower limbs (MOSLL) such as hip fracture surgery(HFS), major knee surgery (MKS) or hip replacement surgery (HRS).

The once-daily recommended dose is 2.5 mgadministered post-operatively by subcutaneous injection.In patients with impaired renal function, a 1.5 mg dose may be considered and in severe cases isrecommended (see "Removal of the contraindication in severe renal impairment" section). The initialdose should be given 6 hours following surgical closure provided that haemostasis has beenestablished. Timing of the first Arixtra injection requires strict adherence in special populations suchas patients 75 years, and/or with body weight

7/31/2019 Wc 500027737

2/45

2/45 EMEA 2004

Solutions with the same excipients but of different concentrations have been used during the clinicaltrials. A bioequivalence study comparing high and low concentrations has been carried out, and nosignificant difference has been observed.

Active substance

Fondaparinux sodium is a sulphated pentasaccharide obtained by total chemical synthesis. It is a whiteto almost white hygroscopic powderand is freely soluble in water, sodium chloride and sodiumhydroxide solutions, and practically insoluble in ethanol. Fondaparinux sodium in the solid state is notlight-sensitive.

The Applicant has confirmed that there is no TSE risk related to the manufacture of the activesubstance or the medicinal product.

Active Substance Specification

The specification includes relevant tests for identity assay, related impurities specific optical rotation,total viable microbial count, endotoxins, appearance of solution, pH, heavy metals, water content,residual solvents, etc. all performed by PhEur methods or by validated in-house methods whereappropriate.

Data on 13 batches indicated satisfactory uniformity and compliance with the specification.

Stability:In addition to short-term stress studies (including photostability studies), long-term data weregenerated from six batches at 25C/60%RH, 30C/60%RH and 40C/75%RH for 24, 12 and 6 monthsof duration respectively. No significant degradation was seen at any of the conditions studied(25C/60%RH, 30C/60%RH and 40C/75%RH). A retest period of two years was accepted at thetime of the authorisation. The retest period has been extended from two to three years further to theassessment of a type I variation.

Other ingredients

The other ingredients of the formulation, hydrochloric acid, sodium hydroxide, and water for injections all comply with PhEur monographs. There are no ingredients of animal origin..

Product development and finished product

The aim of the development work has been to obtain an injectable sterile apyrogenic dosage form (for subcutaneous administration) of fondaparinux sodium.

The same manufacturing process as the commercial one, except for the terminal sterilisation step, has been used during clinical trials. The drug substance is readily dissolved in sodium chloride and water.Compatibility between the solution and the manufacturing equipment has been demonstrated. The

documentation on product development is considered acceptable.The manufacturing process with in-process controls was described in detail and a flow chart provided.Both strengths are manufactured using the same process; the only diference is the fill volume. The

process comprises compounding, sterile filtration with nitrogen gas as a filtration aid, aseptic filling,stoppering, terminal sterilisation inspection, packaging and labelling. Appropriate in-process controlsare performed.

A total validation plan has been provided, describing the confirmatory validation studies for themanufacturing process and in-process control tests.

7/31/2019 Wc 500027737

3/45

3/45 EMEA 2004

Product Specification

The product specification is essentially the same as the active substance specification, with additionaltests for degradation products arising from the manufacturing process and storage over time. Inaddition, product-related tests include particulate matter, extractable volume and sterility.

Data on 14 batches indicate satisfactory uniformity and compliance with the specification.

The development of analytical methods applied to the finished product is fully described; reference ismade, where possible, to PhEur methods.

Stability of the Product

Arixtra 2.5 mg/0.5 mlPhotostability studies have been carried out on two batches of the product. There is no evidence thatthe product is light sensitive and it can be concluded that the product is stable to intense light for 15days.

The product was unaffected by drastic changes in temperature through temperature cycling studies.

At the time of initial submission, long term and accelerated studies under ICH conditions had beencarried out on a large number of batches for up to 12 months. Updated stability data, up to 24 monthsfrom storage at 25C/60%RH, and up to 24 months at 30C/60%RH, were subsequently provided. Ashelf life of two years was accepted. As a consequence of a type I variation, the shelf life was later extended to 3 years. The storage conditions as defined in the SPC remained the same.

Arixtra 1.5mg/03 mlPhotostability studies have been performed in accordance with ICH recommendations (1.2 million luxhours during 24 hours) on one drug product batch. The results are consistent with those for the higher strength and show that the drug product is not sensitive to light and no restrictions are needed in thestorage conditions with respect to light.

Syringes have also been exposed to three cycles at 18C/40C (at least 48 h at each of thetemperatures) and the results indicated that there was no change in the degradation profile or in theother parameters tested.Three batches have been placed on stability studies under ICH conditions. The obtained results are inagreement with those for the higher strength. As suggested by the results of the stability studies and

based on the fact that the only difference between the two strengths is the fill volume, the proposedshelf life for the lower strength product under the conditions specified in the SPC is acceptable. Thestability studies for the applied lower strength will however continue for up to 36 months.

Discussion on chemical, pharmaceutical and biological aspects

The active substance is notable for its complicated synthesis and the problems encountered in thedetection and quantification of related impurities due to the low UV-absorbance of a carbohydratemolecule with no aromatic residues. Finished product manufacture has been evaluated in depth,

particularly with regard to the validation of the process. Product stability has been established on the basis of a large amount of data and comprehensive statistical analysis. Following a list of questions putto the applicant and an assessment of the responses all important quality issues were resolved. Theimportant quality characteristics of the active substance and product are well-defined and controlled,and the product is formulated, manufactured and controlled in a way that is characteristic of a solutionfor injection.

7/31/2019 Wc 500027737

4/45

7/31/2019 Wc 500027737

5/45

5/45 EMEA 2004

observed increase in clearance, which, in turn, was considered to be due to increased free fraction of fondaparinux, resulting from saturation of plasma protein (e.g. ATIII). The systemic exposure tofondaparinux in rat and monkey significantly increased following repeated administration, similarly towhat was observed in man. At upper non-saturating exposure (0.4 mg/kg), drug-related radioactivityin rats receiving [ 35S]-fondaparinux was detected in many specimensfondaparinux. The samples withthe longest terminal half-lives were Harders gland, liver, eye (vitreous humor), bone marrow andkidneys with 10-20 h, and cartilage with >40 h. Retention of drug-related radioactivity was consideredto take place in kidney cortex and cartilage. Drug-related radioactivity was also detected in rat foetusand in breast milk. No radioactive metabolites could be detected in body fluids (blood and urine) or inin vitro tests for potential liver metabolism after exposure to [ 35S]-fondaparinux when assayed withHPLC. Evidence of desulphation metabolism was detected in blood, urine, bile and hepatocytes bymeans of the detection of [ 35S]-SO 4

2-ions. Urine secretion data in rat indicated some activity-reducingmetabolism.The elimination of fondaparinux from the circulation was significantly faster in rat, rabbit and monkeycompared with man. The terminal half-life did not change with dose in the animals. Drug-relatedradioactivity was mainly excreted via urine, whereas the faecal excretion was low (rat) or negligible(monkey).

Toxicology

A large number of studies were performed in Sprague Dawley and Wistar rats and in Cynomolgus monkeys with maximal treatment duration of 3 months. The studies were generally adequately

performed but inclusion of a comparator like a UFH or a LMWH would have been useful in theoutcome assessment. At 10 mg/kg/day, the exposure ratios were approximately 5 (rat) and 16(monkey), compared with the intended clinical exposure.

A single dose of fondaparinux was well tolerated in mouse, rat and monkey at doses up to 40 mg/kg.All repeated dose toxicity studies were performed with doses of 0.4, 2.0 and 10.0 mg/kg/day(approximately 12, 60 and 300 times the proposed therapeutic dose). Few non-haemorrhage relatedeffects were observed. The true toxicity limits of fondaparinux are clearly beyond the chosen dosesand probably differ for different animal species and treatment duration. Supporting arguments for doseselection were provided, including dose limiting effects such as adverse haemorrhagic events (AHE)(monkey) and saturable plasma protein binding (mainly rat). Thus, a significant increase in dosewould be required to achieve a moderate increase in total exposure. This would lead to an increase of the free fraction that would be subject to a fast renal clearance.

Rat: Haemorrhagic, not fully dose-related changes were localised mainly to the injection sites andoccasionally to the draining lymph nodes in the repeated dose toxicity studies. There were no major increases in other haemorrhagic adverse events. Reversible slight decreases in haemoglobin levels,reticulocytosis and an increase of urine urobilinogen, related to bleeding/haematoma-associatedanaemia and erythropoiesis, were observed in some studies. Slight plasma electrolyte level variationswere noted in the 2-week studies and also in the 3-month monkey study.

Monkey: The incidences of injection site haemorrhages and other trauma-related haemorrhagic eventswere increased in a dose-dependent manner. The reversibility of these findings after repeatedadministration of fondaparinux was assessed in one animal/sex from the control and high-dose groupsin the 4-week study, which were followed after treatment. Histopathological treatment-related effects

were not seen in the kidney, the major route of excretion, or in the liver. There were no, or onlyisolated increases in transaminase activity. Large or very large haematomas with or without associatedresolutive inflammation were observed at handling sites, blood puncture sites and self-inflicted traumasites in some treated animals at all doses. Such AHE were associated with a regenerative anaemia.

In the 3-month and in the 4-week monkey studies, the male animals displayed changes to genitalorgans which were related to degree of maturity or to health deterioration.

In the 3-month study, a dose-dependent increase in absolute (male) and relative (male and female) pituitary weights was noted, while the male monkeys had a decrease in T3 hormone at the end of the

7/31/2019 Wc 500027737

6/45

6/45 EMEA 2004

study. However, the observed variation was within the physiological range as compared to the 2- and4-week control monkeys.

No analysis specific for bone tissue (e.g. serum bone specific alkaline phosphatase) or morphometricanalysis was used, but conventional total alkaline phosphatase and bone histology revealed nonegative effects on bone in the rat and monkey.

Impurities During the impurity qualification, an increased incidence of retinal atrophy was observed inthe treated rats in two 2-week studies. This was not considered to be due to impurities, but explainedas a direct phototoxic effect produced by excessive exposure of incandescent light during the infusion

procedure, when, to favour vasodilatation, light was used to warm the animal. This warming procedure was performed only at the site where the atrophies were found and yielded energies thatwere within the reported levels that may induce retinal atrophy.From a toxicological point of view, all impurities are considered qualified.

Reproductive and development toxicity All reproduction toxicity studies were performed with doses upto 10 mg/kg. In the rabbit studies, at the high-dose level, an exposure ratio of 12.6 was achievedcompared to the human. Fondaparinux did not show any major effects on mating performance andfertility, gestation and parturition, lactation, new-born viability and growth, F1 behaviour andreproduction and F2 foetal development. An increased dose-dependent preimplantation loss wasobserved in both rabbit studies. This was explained to be within the physiological range whencomparing in-house and external historical data. Such data should be interpreted with caution andconcurrent control animals are considered more relevant. Moreover, a pharmacological effect couldnot be excluded since one action of sulphated polysaccharides is to inhibit cellular adhesion, asdescribed in a number of publications. The clinical relevance of this finding was not clear, but was acause for concern. The inadequate exposure of the animals (half-life of fondaparinux in human is 17hours as compared to the 1-2 hours in rabbit) further strengthened this concern. However, data werere-calculated by the Applicant and does with high or low numbers of corpora lutea were excluded, asseems reasonable, since literature supports an increased loss in this population. This resulted in a nondose-related response in preimplantation losses. Further, the Applicant submitted convincingarguments that no effect of fondaparinux was likely on cell adhesion and has agreed to addinformation regarding a limited exposure in the SPC, section 4.6.

Genotoxicity The mutagenic potential was evaluated in a battery of tests including bacterial reversegene mutation (Ames) tests, mouse lymphoma gene mutation tests, mammalian in vitro DNA repair test, human lymphocytes chromosome aberration tests and one in vivo rat micronucleus test.Fondaparinux was not mutagenic or clastogenic.

Carcinogenicity No studies were performed since the intended treatment duration in humans is lessthan one month. The chemical structure of fondaparinux does not indicate a possible carcinogenic

potential.

Local tolerance testing was performed during all single, repeated and reproduction toxicity studies. Asexpected, injection site haemorrhages, occasionally associated with an inflammatory infiltration, werenoted.

Immunotoxicity Fondaparinux did not induce any active or passive anaphylactic signs in guinea pigs or rats. No relevant effects were seen on rat bone marrow cellularity or monkey T4, T8, NK and B

peripheral blood lymphocyte counts, peripheral blood lymphoblast transformation tests with PHA,Con A and PWM, peripheral NK cells cytotoxicity test, mixed lymphocyte reaction and total plasmaIgGs and IgMs levels.

Ecotoxicity/Environmental Risk Assessment A brief evaluation on the risk of use, excretion anddisposal of fondaparinux has been conducted. No significant environmental effect is expected.

Discussion on toxico-pharmacological aspects

7/31/2019 Wc 500027737

7/45

7/45 EMEA 2004

Preclinical data reveal no special risk for humans based on conventional studies of safety pharmacology, repeated dose toxicity, and genotoxicity. Animal studies are insufficient with respect toeffects on toxicity to reproduction because of limited exposure.This information is reflected in the approved Summary of Product Characteristics

Part IV: Clinical aspectsClinical pharmacology

Pharmacodynamics

Fondaparinux binds to human ATIII resulting in a 1:1 molar complex with a Kd of 58 nM. Plasmaconcentrations of 4 to 8 mg/l should saturate physiological concentrations of ATIII (2-4 M).Circulating anti-Xa activity was proportional to fondaparinux dose after single or repeated q.d. s.c.dosing in healthy subjects and after repeated q.d. s.c.dosing (0.75 to 8 mg) in patients undergoingmajor orthopaedic surgery.

Fondaparinux has no ATIII-mediated anti-IIa activity except at concentrations beyond ATIIIsaturation, nor does it show significant anti-IIa activity through heparin cofactor II (HCII) activation.

No circulating anti-IIa activity was detected after 0.7-18 mg s.c. or 2-16 mg i.v. Inhibition of thrombingeneration is mediated by the anti-Xa activity. Indirect evidence of in vivo inhibition of thrombingeneration was obtained by assessing thrombin-antithrombin complexes, prothrombin fragments 1+2and D-dimers in healthy volunteers, and in different patient groups.The reduction of thrombin generation also reduces the pro-coagulative feedback exerted by thrombin.This probably contributes to the minor inhibitory effects of fondaparinux on factor IXa and on factor X activation observed.

Fondaparinux does not alter prothrombin time and only slightly prolongs aPTT (maximal mean prolongation 20%) after 0.35-30 mg s.c., and 2-20 mg i.v. The aPTT was doubled only at plasmaconcentrations above 73 mg/l, whereas it is doubled at 2.5 mg/l UFH. Activated clotting time (ACT)has been measured only in patients undergoing PTCA after fondaparinux 12 mg IV. Inconsistent ACTincreases were observed 2 and 48 hours after dosing, but not at 24 hours. These changes were much

smaller than after UFH in this indication.

Unlike UFH, fondaparinux does not interact with ADP or collagen-induced platelet aggregation atconcentrations up to 150 mg/l, and does not bind PF4 or stimulate platelet serotonin release in the

presence of sera from heparin-induced thrombocytopenia (HIT) patients. This argues against a potential to induce HIT, which results from heparin binding to PF4 followed by antibody interactionsand platelet activation. Fondaparinux showed no ex vivo effect on platelet aggregation induced byADP, thrombin, collagen, or arachidonic acid, and did not change in vivo platelet counts after singleand repeated dosing. Fondaparinux does not prolong bleeding time up to daily doses of 18 mg s.c. or 20 mg i.v.

Single or repeated doses of fondaparinux did not affect the ATIII level, but a decrease in ATIII wasobserved in patients undergoing total hip replacement (THR). The decrease was not different betweenthe fondaparinux and enoxaparin treated patients.

One issue that was raised during the assessment was whether there is a risk that the fondaparinuxconcentrations exceed the binding capacity of ATIII in the postoperative phase. The mean ATIIIreductions seen postoperatively after orthopaedic surgery were 10-15% and they were independent of treatment (fondaparinux or enoxaparin). It can be calculated that variations in ATIII levels within awide range affect concentrations of bound fondaparinux only to a small extent. The changes of unbound fondaparinux with varying ATIII levels can be expected to be relatively larger. Theconcentrations of free fondaparinux that can be expected with the recommended dose, even with verylow ATIII levels, are of no major safety concern according to the preclinical and clinical experience.

7/31/2019 Wc 500027737

8/45

8/45 EMEA 2004

The small and transient aPTT prolongation after 10 mg sc fondaparinux was not affected by the co-administration of aspirin (single oral 975 mg dose). The lack of interaction with warfarin on PT andaPTT was documented by a single cross-over study involving 12 subjects. Whether a

pharmacodynamic interaction between fondaparinux and vitamin K antagonists could result in anunexpected bleeding tendency was discussed. Based on clinical experience and on what is knownabout the pharmacodynamic interactions it was concluded that the expected bleeding tendency should

not be significantly different with the combination fondaparinux/warfarin from what is seen withheparin/warfarin.

Pharmacokinetics

The human pharmacokinetics was evaluated in 27 clinical trials, including 6 Phase I studies in healthysubjects, 5 studies in special populations, 5 studies assessing specific drug interactions, and 8 studiesin patients for the current indication. Generally, the pharmacokinetics of fondaparinux were wellcharacterised. The plasma and urine fondaparinux concentrations in humans were assessed byvalidated biological assays (bioassays). These methods utilised the ATIII-mediated factor-Xainhibitory activity of the drug with spectrophotometric detection of a specific factor-Xa substrate.Pharmacokinetic parameters were obtained following non-compartmental and compartmental(population) analysis.

Fondaparinux was rapidly absorbed following s.c injection (2.5mg) and C max of 0.34 0.04 mg/l wasreached after 1.7 0.4h in young healthy volunteers. The mean absolute bioavailability estimate was107% (90% CI: 102%, 112%). The volume of distribution was approximately 8.2 1.1l after s.cinjections of 2.5mg. Fondaparinux is highly bound to ATIII (97.0-98.6%) and in vitro studies haveshown that the binding becomes saturated at concentrations above 2 mg/l. The fondaparinuxconcentration-time profile is described by a bi-exponential decline. The terminal half-life estimatesvaried between 13 and 21h and the CL/F ranged from 5.1 7.9 ml/min, with renal excretion being themajor elimination pathway. The renal CL exceeds the filtration CL and, thus, it is postulated thatfondaparinux is actively secreted in the kidney. In healthy volunteers, steady-state was reached after 3-4 days, which is consistent with the half-life estimation. The accumulation ratio at steady-state was1.3. The exposure is decreased non-proportionally at higher doses, due to saturation of protein

binding, but the total concentrations are fairly linear within the therapeutic range. However, there are

indications of saturability in vivo at lower concentrations than shown in in vitro studies.

Similar pharmacokinetic characteristics were observed for the healthy elderly and the patient populations. The inter-patient variability in CL/F and in V/F in the target population was 30-40% and20%, respectively. The influence of various covariates was assessed in the population analysis andonly weight was subsequently included in the final model as linearly increasing CL/F byapproximately 9.3%/10kg. The lower plasma clearance leading to higher plasma levels withdecreasing body weight was associated with a trend to higher major bleeding rates in patients less than50 kg. Although highly correlated with total CL, renal function was not detected as significantlyexplaining the inter-patient variability in drug levels. The influence of important covariates wasevaluated on post-hoc estimates but the predictive capacity of the population model seems limited.Explanation was given in the August 2001 answers.In a single dose study, impaired renal function resulted in a considerable reduction in CL/F (5-fold

reduction in severe renal impairment), no change in distribution volume and, consequently, prolongedhalf-life. In Phase III trials, average CL/F was 1.3-1.6 fold lower in patients with mild renalimpairment and 1.6-2.3 fold lower in patients with moderate renal impairment, compared with patientswith normal renal function. Arixtra is contraindicated in patients with very severe renal impairment(CL crea< 20 ml/min) and the use of a lower dose is recommended in patients with severe renalimpairment (CL crea between 20 and 30 ml/min) and may be considered in patients with moderate renalimpairment (CL crea between 30 and 50 ml/min) . No pharmacokinetic evaluation was performed withrespect to hepatic impairment, since the contribution of hepatic elimination to the overall eliminationwas assumed to be limited, which is acceptable. Studies in elderly patients showed an expectedcorrelation between age and CL/F most likely explained by reduced renal function in the elderly.

7/31/2019 Wc 500027737

9/45

9/45 EMEA 2004

Taking demographic differences into account, the pharmacokinetic profiles were similar in men andwomen.

The relation between plasma concentrations (C max) and VTE or major bleedings (MB) was evaluatedin DRI2643 (dose range 0.75mg to 8mg). Patients with MB tended to have higher mean C maxss plasmaconcentration than patients without events, and patients with VTE tended to have lower mean C maxss

plasma concentration than patients without events, confirming the dose-response data.

There was a consistent trend to higher MB rate with increasing degree of renal impairment, withincreasing age, and with decreasing body weight. It was demonstrated that the incidence of MB washigher in patients receiving the first post-operative dose early (

7/31/2019 Wc 500027737

10/45

10/45 EMEA 2004

Number of centresand of Randomizedpatients

Efficacy assessment

DRI2643Canada,70 centres

N=950

Total hip replacement (THR)Double blindAssessor blind

0.75, 1.5, 3, 6, 8 mg s.c.started 6 2 h postoperativelyComparator:Enoxaparin 30mg x 2

Venography day 5-10

95001USA19 centres

N=318

Total knee replacement(TKR)OpenAssessor blind

0.75, 1.5, 3, 6, 8 mg s.c.started 6 2 h postoperatively

noncontrolled

Venography day 5-10

The number of patients with DVT found at the venographic assessment is given in tables 2 and 3.

Table 2 -Number (%) of Patients with any DVT, any Proximal DVT, and Distal only DVTDuring the Treatment Period in DRI2643

FONDAPARINUX Enoxaparin

DVT 0.75 mg

(N = 102)

1.5 mg

(N = 101)

3.0 mg

(N = 101)

6.0 mg

(N = 44)

8.0 mg

(N = 22) (N = 150)

Any DVT n (%) 11 (10.8) 6 (5.9) 2 (2.0) 2 (4.5) 0 (0.0) 14 (9.3)

Any proximalDVT

n (%) 3 (2.9) 5 (5.0) 1 (1.0) 1 (2.3) 0 (0.0) 5 (3.3)

Distal only DVT n (%) 8 (7.8) 1 (1.0) 1 (1.0) 1 (2.3) 0 (0.0) 9 (6.0)

Table 3 - Number (%) of Patients with any DVT, any Proximal DVT, and Distal only DVTDuring the Treatment Period in 95001

FONDAPARINUX

DVT 0.75 mg o.d.

(N = 22)

1.5 mg o.d.

(N = 44)

3.0 mg o.d.

(N = 44)

6.0 mg o.d.

(N = 18)

8.0 mg o.d.

(N = 13)

Any DVT n (%) 13 (59.1) 15 (34.1) 6 (13.6) 2 (11.1) 2 (15.4)

Any proximal DVT n (%) 1 (4.5) 2 (4.5) 0 (0.0) 1 (5.6) 0 (0.0)

Distal only DVT n (%) 12 (54.5) 13 (29.5) 6 (13.6) 1 (5.6) 2 (15.4)

The number of MB is given in Table 4. An adjudication committee blindly classified the bleedingreports. The definition of MB was essentially the same as used in the phase III studies. Due toincreased bleeding the study arms of 6 and 8 mg were discontinued prematurely.

7/31/2019 Wc 500027737

11/45

7/31/2019 Wc 500027737

12/45

12/45 EMEA 2004

age less than 18 years serum creatinine level over 180 mol/l (2.0 mg/dl) known bleeding tendency recent stroke thrombocytopenia or a history of thrombocytopenia (

7/31/2019 Wc 500027737

13/45

13/45 EMEA 2004

45/3608 (1.2%) ]. Thus, no apparent reduction of symptomatic VTE by fondaparinux treatmentcorresponding to the reduction of asymtomatic DVT could be seen.

The CPMP accepted that a reasonable explanation for the nominal difference regarding incidence of symptomatic VTE was the following: About 6.2 % of the evaluable patients in the fondaparinux groupreceived curative treatment after the primary endpoint evaluation (phlebography) as compared with11.9 % in the enoxaparin group. This type of difference would be expected in a situation when thecompared products have different activity for prevention of DVT and when the large majority of

endpoint events consists of venographically diagnosed, asymptomatic DVT, which, nevertheless,trigger curative treatment in a large proportion of patients. The bias introduced by leaving a larger fraction of patients at some continuing risk for VTE in the fondaparinux group without anticoagulanttherapy for the remainder of the study period would also be expected to result in a somewhat higher incidence of symptomatic VTE in the fondaparinux group after the venographic evaluation.

For patients treated with fondaparinux or enoxaparin but not evaluable concerning the primaryefficacy end-point (no venography performed), the incidences of symptomatic VTE between day 12and day 49 did not differ between treatment groups (fondaparinux 1.3%, enoxaparin 1.0%).

The composite endpoint chosen is somewhat different from the endpoint recommended in the Pointsto consider document (CPMP/EWP/707/98) that includes all-cause mortality. However, the endpointwas chosen in accordance with scientific advice given by the CPMP in 1997. The overall conclusions

regarding efficacy and safety remained unchanged after a reanalysis made retrospectively andincluding all-cause mortality.An ITT analysis (a best case scenario) including all patients that went through surgery and had anystudy medication administered gave additional support for the conclusions made on the overallefficacy of fondaparinux.

Clinical safety in MOSLL

Bleeding Most bleedings occurred at the surgical sites. No fatal, CNS or retroperitoneal bleedings were reportedin the fondaparinux 2.5 mg group.

In the two dose-ranging studies there was a significant dose response for MB as well as for all

bleeding events and the percentage of transfused patients.

An adjudication committee blindly classified the bleeding reports in all the phase III studies. Onlystudies from the same randomisation type were pooled for integrated analyses since intraoperative

bleeding was included only in the studies with preoperative randomisation. The frequencies of bleedings are given in Table 9.

Table 9. Number (%) of patients (Fondaparinux/ Enoxaparin) experiencing Bleedings up to day11.Preoperative randomisation*Study, surgery N Major Bleedings Minor bleedings onlyEphesus, THR 1140/1133 47(4.1%)/32(2.8%) 44(3.9%)/38(3.4%)Penthifra, HFS 831/842 18(2.2%)/19(2.3%) 34(4.1%)/18(2.1%)Pooled Results 1971/1975 65(3.3%)/51(2.6%) 78(4.0%)/56(2.8%)* enoxaparin European dosing regimen = 40 mg q.d

Postoperative randomisation*Study, surgery N Major Bleedings Minor bleedings onlyPentathlon 2000, THR 1128/1129 20(1.8%)/11(1.0%) 17(1.5%)/24(2.1%)Pentamaks, MKS 517/517 11(2.1%)/1(0.2%) 14(2.7%)/19(3.7%)Pooled Results 1645/1646 31(1.9%)/12(0.7%) 31(1.9%)/43(2.6%)* enoxaparin North- American dosing regimen = 30 mg bid

7/31/2019 Wc 500027737

14/45

14/45 EMEA 2004

An overall tendency, although not significant, for higher major bleeding rates in the fondaparinuxgroups can be seen. Only a fraction of all minor bleedings can be expected to have been reported.

The percentage of patients that had a decrease in haemoglobin concentrations 2 g/dl during thetreatment period was higher in the fondaparinux treated group as compared with the enoxaparin group,

both in studies with preoperative randomisation (22.0% vs.16.0%) and postoperative randomisation(55.5% vs. 49.6%). The rates of bleedings related to some of the base-line covariates are given inTable 10.

Table 10. Number (%) of Patients Experiencing Major Bleeding Events From FirstFondaparinux Injection up to Day 11 by Baseline Covariates - All fondaparinux 2.5 mg TreatedPatients in Orthopaedic Surgery Studies

fondaparinux 2.5 mg(N = 3595

Type of surgery Total Hip Replacement 51/2249 (2.3%)Hip Fracture 18/829 (2.2%)Total Knee Replacement 11/517 (2.1%)

Gender Male 34/1429 (2.4%)Female 46/2166 (2.1%)

Age (years) Missing 0/4 (0.0%)

7/31/2019 Wc 500027737

15/45

15/45 EMEA 2004

More bleedings were seen in patients that got the first fondaparinux injection soon after surgery (< 6hours). Again, this tendency was more pronounced in patients of increased age, with low body-weightand with impaired renal function. Strict adherence to the recommended timing of the first Arixtrainjection reduced the risk of MB among the fragile patients to the same level as normal patients . For patients treated according to the finally proposed regimen (first dose 6 hours following surgicalclosure in all types of major orthopedic surgeries of the lower limbs), the incidences of MB were 2.8%and 2.6% in the fondaparinux and enoxaparin 40 mg groups, respectively and this was reflected in theSPC section 5.1.

Concerning race no conclusions were possible due to the low number of non-Caucasians included.There was no effect of surgery type or of type of anaesthesia on MB rates.

Other adverse events and serious adverse events/deaths

The mortality rates were low and did not differ between the treatment groups (Table 11).

Table 11. Number of deaths during treatment (until day 11) and study (until day 49) periods.Deaths until day 11 Deaths until day 49

fondaparinux Enoxaparin fondaparinux EnoxaparinPreoperative randomisation 11/1971 18/2051 40/1971 46/2051Postoperative randomisation 4/1645 3/1906 8/1645 7/1906

Deaths due to bleeding, all studies 0 1 0 2Deaths due to PE, all studies 2 3 11 11

Apart from the deaths caused by bleeding or PE, none of the deaths was suspected to be associatedwith the treatment by the investigators. The higher mortality rates in the studies with preoperativerandomisation are due to the higher mortality in the HFS study (Penthifra), including elderly, acutelyill patients.Concerning other adverse events such as surgical site reactions, myocardial infarctions, cardiac failure,cerebrovascular disorders, no differences between treatment groups were seen in the 4 phase IIIstudies. The rate and spectrum of reported adverse events not related to bleeding did not deviate fromwhat could be expected. No specific suspected drug-related other adverse events were reported.

Laboratory findings

In the clinical studies the percentages of patients experiencing a decrease in platelet count to < 100 x109/l were not related to fondaparinux dose and were similar to those observed in the enoxaparingroup.In the orthopaedic surgery studies a slightly higher rate of patients developed positive ELISA tests inthe preoperatively than the postoperatively randomised studied. No differences were seen between thetreatment groups. The higher number of positive tests associated with positive serotonin release testsin the preoperatively randomised studies cannot be explained. The development of a positive test didnot appear to be predictive of thrombocytopenia or of thrombotic event.

No thrombocytopenias with suspected immunoallergic pathophysiology (HIT, type II) weredocumented in the clinical development program.Taking in to account that fondaparinux lacks PF4 binding capacity fondaparinux did not elicit any 14C-serotonin release in the presence of sera from HIT patients and

platelets from healthy subjectsthe risk of thrombocytopenia of immunoallergic origin (HIT type II) induced by fondaparinux can beexpected to be low from available documentation. However, until further experience is gained, awarning has been included in the SPC that fondaparinux should not be used in patients with a historyof HIT type II.Thrombocytopenia cases will be carefully monitored and reported by the Marketing AuthorisationHolder in the Periodic Safety Update Reports.

A transitory increase in transaminases is observed during treatment with heparins. In the DVTtreatment study (DRI2440) no significant mean increase in AST or ALT was observed under

7/31/2019 Wc 500027737

16/45

16/45 EMEA 2004

fondaparinux in contrast to dalteparin, with consistent observations of larger percentage of patientsexperiencing increases above the ULN. No increases were seen after fondaparinux in patients under-going PTCA or hemodialysis (studies ACT2445, 63113).There was no difference in transaminase levels between fondaparinux and placebo in the phase Istudies and no dose response in the phase II studies. As compared with the enoxaparin treated group inthe pooled patient material in orthopaedic surgery the mean increase in AST, ALT levels wassignificantly less pronounced among the fondaparinux-treated patients and significantly fewer patientsexperienced elevations above ULN or 3xULN. The increases seen in the fondaparinux treated patientsmay be related to surgery and physiological reactions in the post-operative phase rather than tofondaparinux .

Safety in special populations

The rate of MB was inversely related to renal function and a tendency for an inverse relation was alsoseen with body weight. The rate of MB tended to be increased with age.

The safety in fragile patients (elderly, low body weight, impaired renal function) was a concern.Bleeding tendency in these patients may to a large extent be explained by co-morbid factors, but itremains that, with the proposed posology, there will, inevitably, be accumulation of fondaparinux andexposure considerably in excess of that seen in patients with normal excretory capacity.By the strict adherence to the recommended timing of the first Arixtra injection, as emphasised by theapplicant, the risk for MB may be reduced.Renal failure should be expected to be associated with prolonged exposure to higher plasmaconcentrations of fondaparinux apart from the haemostatic disturbances caused by the kidney failure

per se . Fondaparinux concentrations are significantly increased in severe renal impairment (CL cr :75 years old,

7/31/2019 Wc 500027737

17/45

17/45 EMEA 2004

0

1

2

3

4

5

6

7

8

< 4hours

[4,5[ [5,6[ [6,7[ [7,8] ]8,9] ]9,10] >10hours

Time Of First Post OPerative Injection

% o

f M a j o r

B l e e d

i n g

Not Fragile

Fragile2/61

3/45

5/307

9/182

8/452

8/290

17/877

11/526 9/3984/211

0/76 0/31 0/16 0/7

1/56

1/32

Fragile Patients :age 75 yearsand/or body weight < 50 kg and/or CLcr 10hours

Time Of First Post OPerative Injection

% o

f M a j o r

B l e e d

i n g

Not Fragile

Fragile2/61

3/45

5/307

9/182

8/452

8/290

17/877

11/526 9/3984/211

0/76 0/31 0/16 0/7

1/56

1/32

Fragile Patients :age 75 yearsand/or body weight < 50 kg and/or CLcr

7/31/2019 Wc 500027737

18/45

18/45 EMEA 2004

The dose, dose regimen and duration of treatment have been substantiated both in the general population and in the fragile patients groups (low body weight, elderly, moderate and severe renalfailure). The information is reflected in the Summary of Product Characteristics.

Guidance for patients that need prolonged prophylaxis (beyond 9 days) was discussed in depth by theCPMP. Prolonged prophylaxis, beyond 9 days after surgery, was not tested in the pivotal trials of MOSLL hitherto assessed. Thus, a benefit/risk assessment for such prolonged treatment with

fondaparinux could not be made at the time of the initial MA. Therefore, guidance on how to switch to prophylaxis with heparin, LMW-heparin or to treatment with a vitamin K antagonist (VKA) in casefollow-up therapy was deemed necessary beyond the peri-operative period was included in theSummary of Product Characteristics. This guidance has subsequently been deleted further to theassessment of the Penthifra Plus study data (see prolonged prophylaxis section) ., .

Fondaparinux did not affect the prothrombin time in healthy volunteers also given warfarin (study63108). Warfarin did not affect the pharmacokinetics of fondaparinux. In study DRI2440 (doseranging DVT treatment study) over 300 patients have been treated concomitantly with fondaparinux(doses 5, 7.5 or 10 mg) and oral anticoagulants. As compared with dalteparin and oral anticoagulants,the same time of combined treatment was required to reach therapeutic INR values. There was noincreased bleeding tendency in the fondaparinux group as compared to the dalteparin group. Inconclusion there is no suggestion that fondaparinux when combined with oral anticoagulants should

lead to a higher and unexpected bleeding frequency in comparison with heparin or LMWH. Theinformation is included in the Summary of Product Characteristics.

Safety

The three main areas of concern discussed by the CPMP were the bleeding risk (particularly in fragile populations), the potential for interaction with other medicinal products used in prolonged prophylaxis, and the occurrence of thrombocytopenia.

These points have been addressed with appropriate additional analyses of the phase II and phase IIIextensive studies submitted. The SPC has been amended as necessary with contraindication of the useof Arixtra in very severe renal failure patients and introducing the necessary warning and precautionfor use in other populations. The assessment of the pharmacokinetic, interaction and dose response

studies, as well as a specific pooled analysis of all patients treated with other medicinal products for prolonged prophylaxis during the Phase III program has provided reassurance that there are noobvious indications of an unexpected synergistic effect on bleeding tendency following the combinedtreatment of fondaparinux and oral anticoagulants.

The same incidence of thrombocytopenia was observed in the two treatment groups across the phaseIII studies. Taking into account that no documented cases of immunoallergic thrombocytopenia have

been reported across the whole clinical experience accrued so far with Arixtra, and considering thatfondaparinux lacks PF4 binding capacity, the risk of thrombocytopenia of immunoallergic origin (HITtype II) induced by fondaparinux can be expected to be low. However, the CPMP recommended asfollow-up measure that until further experience is gathered, thrombocytopenia cases be carefullymonitored and reported by the Marketing Authorisation Holder in the Periodic Safety Update Reports.

Benefit/risk assessment

The Applicant has provided substantial evidence documenting the efficacy and safety of fondaparinuxwhen used as directed in the Prevention of Venous Thromboembolic Events (VTE) in patientsundergoing major orthopaedic surgery of the lower limbs such as hip fracture, major knee surgery or hip replacement surgery.

Appropriate contra-indication in patients with very severe renal failure, recommendation for theduration of treatment as well as clarifications and undertaking of follow-up on the cases of thrombocytopenia which might occur during treatment with fondaparinux were subjects of in-depthdiscussion with the applicant in the course of an oral explanation with the CPMP.

7/31/2019 Wc 500027737

19/45

19/45 EMEA 2004

Outstanding issues relevant to the complex manufacturing and quality control process of fondaparinuxhave been extensively addressed in the information and clarifications provided by the applicant. Anumber of minor quality points will be further addressed by the applicant after MarketingAuthorisation is granted, with the generation of additional analytical data.

The over-all benefit/risk balance for Arixtra, used as recommended in the SPC, is considered to befavourable.

RecommendationBased on the CPMP review of data on quality, safety and efficacy, the CPMP considered by consensusthat the benefit/risk profile of Arixtra in Prevention of Venous Thromboembolic Events (VTE) in

patients undergoing major orthopaedic surgery of the lower limbs such as hip fracture, major kneesurgery or hip replacement surgery was favourable and therefore recommended the granting of themarketing authorisation.

EXTENDED PROPHYLAXIS IN HIP FRACTURE SURGERY

Rationale

The request for extended prophylaxis is based on the results of the Penthifra Plus study (EFC4582) in patients undergoing Hip Fracture Surgery (HFS). No similar studies appear to have been performed inHFS prior to this pivotal study but long term prophylaxis experience was reported with LMWH inHRS. Indeed, since 1996 at least 6 placebo-controlled studies with LMWHs have demonstrated areduction of venographically detected DVT approximately 1 month after elective HRS if prolonged

prophylaxis (additional prophylaxis for approximately 3 weeks after the first post-operative week) isutilised. In summary, the results from these 6 studies demonstrate a reduction of DVT(venographically detected and symptomatic) from 22.5% in the pooled placebo group to 8.5% in theLMWH group. Proximal DVTs are also reduced from 11.2 to 3.0% and symptomatic VTE from 4.4 to1.4% in the placebo and LMWH groups, respectively. Nevertheless, prolonged prophylaxis withLMWH has not been generally accepted for routine use in patients undergoing HRS. The labelling inseveral countries does, however, allow prolonged prophylaxis in patients deemed to suffer from anincreased thrombotic risk.

Penthifra plus Trial DesignA multicentre, international, randomised, double-blind, placebo-controlled trial of fondaparinux 2.5mg qd for 21 2 days following an initial open-label prophylaxis of 7 1 days, for the prophylaxis of VTE in patients undergoing HFS. Patients without evidence of symptomatic VTE or disqualifying

bleeding during the initial week of prophylaxis and without study exclusion criteria could berandomised into the double-blind study.

The primary efficacy outcome was the rate of VTE (defined as DVT detected by mandatory bilateralvenography, or documented symptomatic DVT, or fatal/non-fatal PE) during the randomised, double-

blind extended prophylaxis period. The mandatory venography was to be performed at the end of theextended prophylaxis period (Day 21 2). The same composite efficacy and safety endpoints andadjudication (Hamilton) used in all confirmatory studies of the original marketing authorisationapplication (MAA) were chosen, facilitating cross-study comparisons. The evaluable population for the primary efficacy analysis was defined as all treated patients with a VTE assessment up to Day 24included. Patients were to be analysed in the group to which they were randomised. In order to explore

possible effects of a missing VTE assessment on the interpretation of the efficacy results, the samesensitivity analyses (best case, realistic case and worst case) as those included in the original MAAwere specified in the protocol. The safety population was the as treated population, with a patient

being considered in the fondaparinux group as soon as (s)he received at least one dose after randomisation. The main safety analysis considered all events observed up to Day 25, i.e., 2 additionaldays after the end of treatment.

7/31/2019 Wc 500027737

20/45

20/45 EMEA 2004

A sample size of 210 patients per group was estimated assuming a rate of VTE after 3 weeks of extended thromboprophylaxis of 11.5%, a risk reduction of 50%, an expected incidence of VTE in the

placebo group of 23% and a power of 85%. Assuming 10% of patients would not enter the double- blind period and 30% would be non-evaluable for primary efficacy, the target number of patients inthe pre-randomised period was 670 in order to reach a total of 600 randomised patients (300/group).

ResultsA total of 737 patients were enrolled, 81 (11%) of which were not subsequently randomised into theextended prophylaxis study. The most common reasons for subsequent non-randomisation were: 39(S)AEs (serious adverse event) or efficacy failure [including 5 investigator-identified DVT/PE, 8

bleedings and 34 (S)AE], and 42 neither efficacy failure nor (S)AE [including 11 withdrawn informedconsents and 31 other exclusion criteria/reasons]. The exposure of patients to fondaparinux beforerandomisation was 7.0 0.8 days (mean SD). Thus, 656 patients were randomised to either fondaparinux or placebo. By Day 21 2 (end of randomised double-blind period) 428 patients(65.2%) were included in the primary efficacy analysis. The as treated population and the evaluable

population for efficacy were evenly distributed between the 2 treatment groups.

The demographic characteristics of patients are consistent with those of the peri-operative HFS study(EFC2698) in the original MAA. As expected for patients randomised into HFS studies they tended to

be older, lighter, female, and the incidence of impaired kidney function tended to be higher than in patients randomised into the studies of other types of MOSLL .

EfficacyThere was a highly significant reduction of the primary efficacy endpoint in the fondaparinuxtreatment arm as compared to the placebo group (Table 1).

Table 1 - Summary of efficacy results and subsequent curative treatment -Patients [n/N (%)] evaluable for the relevant endpoint

Efficacy Endpoint Fondaparinux 2.5 mg qd Placebo p-Value

VTE (Primary efficacy analysis) 3/208 (1.4) 77/220 (35.0) 4 X 10 -22

DVT 3/208 (1.4) 74/218 (33.9) 4 X 10 -21

Proximal DVT 2/221 (0.9) 35/222 (15.8) 3 X 10-9

Symptomatic VTE 1/326 (0.3)

(1 proximal DVT)9/330 (2.7)

(1 fatal PE, 2 non-fatal PE,5 proximal DVT, 1 distal DVT)

0.021

Patients with curative treatment

(primary efficacy population)5/208 (2.4) 50/220 (22.7) 5 X 10 -11

The VTE rate observed in the fondaparinux treatment group was low (1.4%), lower than that observedin study EFC2698 (8.3%) following peri-operative prophylaxis with fondaparinux in the same patient

population. In this respect, the MAH suggests that prolonged prophylaxis not only prevents thedevelopment of new thrombi, but also may provide an environment that allows lysis of thrombiexisting at the end of the peri-operative prophylaxis period. When the overall HFS population was

subgrouped according to whether or not the patient received a hip replacement (either partial or total prosthesis) the observed VTE rates [fondaparinux: 1/56 (1.8%) vs. placebo: 25/70 (35.7%)] wereconsistent with the overall study results. The results of all 3 pre-specified sensitivity analyses of the

primary analysis were also consistent and statistically significant (best case: RRR = 96%; 95%CI[88%; 100%]; realistic case: RRR = 60%; 95%CI [45%; 73%]; worst case: RRR = 35%; 95%CI [21%;47%]).

The secondary efficacy analyses were consistent with the primary analysis; there was a statisticallysignificant reduction in the rate of symptomatic VTE and a significant reduction of the combinedendpoint of proximal DVT and/or PE (fondaparinux: 2/221 vs. placebo: 38/224).

7/31/2019 Wc 500027737

21/45

21/45 EMEA 2004

Table 1 also shows that considerably fewer patients received curative treatment than those classifiedas having VTE (50 vs. 77 in the placebo group). Further to a request from CPMP, the MAH provideddata showing that, unexpectedly, antithrombotic therapy was initiated by the investigators in only asmall minority (3%) of the patients without any centrally adjudicated VTE. Even more unexpectedly,antithrombotic treatment was not initiated by the investigators in 12/37 patients with proximal DVTand in 22/40 with distal DVT (according to central reading). This point is addressed in the discussion.

The symptomatic events observed in the placebo group tended to present a week or more after randomisation, suggesting that it might take some time for symptomatic VTE to develop (Figure 1).

Figure 1 -Day of symptomatic VTE after randomisation

Period of mandatory

venogram i f no prior symptoms

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24

Placebo O O O O O O O O

O

Fondaparinux O

Period of mandatory

venogram i f no prior symptoms

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24

Placebo O O O O O O O O

O

Fondaparinux O

Period of mandatory

venogram i f no prior symptoms

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24

Placebo O O O O O O O O

O

Fondaparinux O

The trend for a delay in the occurrence of symptomatic VTE in the placebo group contributes to theimpression of a considerable efficacy of fondaparinux prophylaxis in the early post-operative phase. Aconsistent trend for effect of fondaparinux was observed in several patient subgroups. The very lownumber of VTE in fondaparinux-treated patients does not allow identification of covariates bytreatment interactions. However, in the placebo group gender, age, and increased kidney dysfunctionappear to be associated with an increased VTE risk - age and female gender have previously beenreported as VTE risk factors in HS, but not RI. The relationship of VTE risk with RI may be related toage and/or may be a marker of underlying concomitant morbidities that are actually responsible for theincreased risk . Generally, ill patients are prone both to VTE and RI and thus, conclusions on a directcausal relationship between RI and VTE are difficult to draw. Hence, the observed increasedthrombotic tendency has to be weighed against the well known increased bleeding risk in a benefit/risk

evaluation of antithrombotic prophylaxis in RI patients.

SafetyThe number of patients who discontinued study drug by reason is given in table 2.

7/31/2019 Wc 500027737

22/45

22/45 EMEA 2004

Table 2 - Number (%) of patients who discontinued study drug prematurely by primary reason for discontinuation - All treated patients (as randomised groups)

Premature Treatment Discontinuation/Reason for StoppingFondaparinux

(N=326)Placebo(N=330)

Patients who discontinued study drug prematurely 40 (12.3%) 46 (13.9%)Reason(s) for discontinuation a

Lack of efficacy 2 ( 0.6%) 12 ( 3.6%)DVT a 1 ( 0.3%) 9 ( 2.7%)PE a 1 ( 0.3%) 3 ( 0.9%)

AE/SAE b 20 ( 6.1%) 14 ( 4.2%)Bleeding AE/SAE 5 ( 1.5%) 1 ( 0.3%)Repeated surgery 1 ( 0.3%) 3 ( 0.9%)Suspicion of drug-induced decrease of platelet count 0 ( 0.0%) 0 ( 0.0%)Other AE/SAE 14 ( 4.3%) 10 ( 3.0%)

Subject withdrawn consent 12 ( 3.7%) 16 ( 4.8%)Other reason 6 ( 1.8%) 4 ( 1.2%)

NOTE: Patients are considered in the treatment group in which they were randomised.a According to the Investigators judgment.

bIncluding AEs recorded before the first double-blind study drug injection (based on the datacollected in the end of treatment form).

There were numerical increases in bleedings (Table 3) and bleeding-related criteria (i.e., transfusion,decrease in haemoglobin plasma level) under fondaparinux as compared to placebo treatment. Thenumerical difference in major bleedings (MB) was not statistically significant and was due to anincrease in bleeding index (BI) 2 (i.e., a need for transfusion and/or a decrease in haemoglobin level).All of the MB occurred at the surgical site.

Table 3 - % of patients with adjudicated bleeding events during EFC2698 and EFC4582 byadjudication criterion -As treated patients

Peri-Operative ProphylaxisEFC2698 (Penthifra) a

Extended ProphylaxisEFC4582 (Penthifra Plus) b

Patients With

Fondaparinux2.5 mg qd(N=831)

Enoxaparin40 mg qd(N=842)

Fondaparinux2.5 mg qd(N=327)

Placebo(N=329)

Major bleeding 18 (2.2%) 19 (2.3%) 8 (2.4%) 2 (0.6%)

Fatal bleeding 0 (0.0%) 1 (0.1%) 0 (0.0%) 0 (0.0%) Non-fatal critical bleeding 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%)Other non-fatal major bleeding 18 (2.2%) 18 (2.1%) 8 (2.4%) 2 (0.6%)

At surgical site leading to re-operation 3 (0.4%) 2 (0.2%) 2 (0.6%) 2 (0.6%)Only bleeding index (BI) 2 15 (1.8%) 16 (1.9%) 6 (1.8%) 0 (0.0%)

Minor bleeding only 34 (4.1%) 18 (2.1%) 5 (1.5%) 2 (0.6%)

Any bleeding 52 (6.3%) 37 (4.4%) 13 (4.0%) 4 (1.2%)a From first double-blind injection up to Day 11 (Day 1 is the day of surgery).

b From first double-blind injection up to Day 25 (Day 1 is the day of first double-blind injection).

There was a temporal trend for MB, with bleedings more frequently observed immediately after randomisation (Figure 2), although 5 (4 fondaparinux vs 1 placebo) of the 10 MB reported after randomisation started before randomisation.

7/31/2019 Wc 500027737

23/45

23/45 EMEA 2004

Figure 2 - Day of MB from the first study drug administration in EFC4582 as treated patients

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24

Placebo O O

Fondaparinux O O O a O O 1 O2a

O a O a

a : patients with MB started before randomization and worsening after randomization1 : Patient 22050019 : MB only related to BI 2 which only led to study drug temporary discontinuation for 48 hours2 : Patient 10020004 : MB started before randomization which led to re-operation on D15 after randomization

D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11 D12 D13 D14 D15 D16 D17 D18 D19 D20 D21 D22 D23 D24

Placebo O O

Fondaparinux O O O a O O 1 O2a

O a O a

a : patients with MB started before randomization and worsening after randomization1 : Patient 22050019 : MB only related to BI 2 which only led to study drug temporary discontinuation for 48 hours2 : Patient 10020004 : MB started before randomization which led to re-operation on D15 after randomization

Because of the overall low number of MB, particularly in the placebo group, it was not possible toevaluate either the patient characteristics that might correlate with increased bleeding risk or theinteraction of fondaparinux with any concomitant diseases or treatments (Table 4). Nevertheless, thetendency for increased MB rates with decreasing body weight and renal function are consistent withthose observed during peri-operative prophylaxis in the original MAA. The safety concerns for

patients with RI remain and must be carefully considered (see Removal of contraindication in severerenal impairment).

Table 4 - Number (%) [95% CI] of patients with MB by selected covariate - as treatedpopulation

Fondaparinux2.5 mg qd(N=327)

Placebo(N=329)

Gender

Male 2/92 (2.2) ; [0.3;7.6] 1/98 (1.0) ; [0.0;5.6]

Female 6/235 (2.6) ; [0.9;5.5] 1/231 (0.4) ; [0.0;2.4]

Age (years)

7/31/2019 Wc 500027737

24/45

24/45 EMEA 2004

Immune thrombocytopenia and an increase in transaminases are ADRs often associated with heparins.Clinical laboratory evaluations revealed that no patient experienced a platelet count

7/31/2019 Wc 500027737

25/45

25/45 EMEA 2004

agents, as was done in the original MAA. It is also well understood that once a DVT has beendiagnosed at a prescheduled venography it cannot be left untreated in the majority of cases. However,this generally applied rule for clinical decision-making cannot be regarded as scientific proof for theclinical importance of non-symptomatic DVTs 4 weeks after HFS.

As regards extrapolating the Penthifra trial results to HRS patients, the MAH argues that given thealready documented persistence of VTE risk in patients undergoing HRS and the demonstratedefficacy of a less effective agent (enoxaparin), the efficacy of fondaparinux in preventing persistingVTE risk in both types of hip surgery (HRS and HFS) is unquestionable. While the CPMP concur that

prolonged prophylaxis in HRS can be expected to be effective to an extent that is similar (although notnecessarily identical) to the demonstrated efficacy in HFS, the mean incidence of symptomatic VTEafter day 11 was, in relative terms, 27% lower in HRS as compared to HFS. The extrapolation is thusnot endorsed, as reflected in the adopted SPC.

Based on the above reasoning, the CPMP does not believe that a general recommendation for routinely applied prolonged prophylaxis in all patients undergoing HFS is justified. Nevertheless, theresults of this study support a change in the labelling allowing prolonged prophylaxis in patients whoare deemed by the treating physician to suffer from an increased thrombotic risk. This is reflected inthe SPC.

Safety:The number of patients who discontinued treatment prematurely was similar in both treatment groups,with more symptomatic VTE in the placebo group and a tendency for more bleedings in the activelytreated group. As suggested from the results of the studies of perioperative prohylaxis in the originalMAA, a temporal association between time of surgery and the incidence of MB was also observed inthe Penthifra plus trial, although no trend for increased bleeding with increased treatment time is seen,which is interpreted by the MAH as an indication of a lack of a clinically relevant accumulation of fondaparinuxin plasma during the 3 weeks extended treatment. However, considering i) the well-known difficulties in evaluating bleeding tendency in clinical studies, ii) the limited size of the study,iii) the lack of external support for the safety of fondaparinux prophylaxis longer than 9 days, iv) theexclusion of patients with increased bleeding risk and severe RI and v) the avoidance of concomitantmedication interfering with the haemostatic mechanisms in the clinical trial setting which seems tohave been well controlled, it is difficult to conclude from these findings on the risk of bleedings inclinical routine and on the claimed lack of clinical relevance of possible plasma accumulation of fondaparinux (plasma levels were not measured in this study). Plasma concentrations can be viewed asa relevant marker for the safety (and efficacy) of fondaparinux.

REMOVAL OF THE CONTRAINDICATION IN SEVERE RENAL IMPAIRMENT

Scientific background Arixtra was initially contraindicated in patients with severe renal impairment (CL crea

7/31/2019 Wc 500027737

26/45

26/45 EMEA 2004

and normal RI were not included in the analysis and data from the Phase II/III studies were not used,although they included many patients with moderate RI and a few patients with CL crea 20 - 30 ml/min.

Further to a request from CPMP, the population analysis was redone. A population pharmacokinetic(POPPK) model using all available sparse data by SC route from the phase II/III studies was built with

NONMEM software (FOCE interaction) and implemented on Pharsight Trial Simulator (TS2)software. The objective of this new analysis was to build a POPPK model including at least CL crea ascovariate of fondaparinux CL in patients undergoing MOSLL. A total of 756 patients (3413 time

points) with an available CL crea ranging from 23.5 to 231.5 ml/min (550 from study DRI2643, 65 fromstudy ECF2442 and 141 from study EFC2698) were used to develop this model. A 2-compartmentmodel with a first order absorption and elimination from the central compartment best fitted availabledata. Two covariates were included in the final model: CL crea on CL, and the body weight (WGT) oncentral compartment volume (V2).

Table 5 Description of final population pharmacokinetic modelParameter Population

EstimateStandard Error Coefficient of

Variation (%)InterindividualVariability (%)

CL (L/h) 0.0409 0.0160 39 24

CL cr (CL) 0.320 0.0169 5 -

CL(CLcr) = 0.0409 + 0.320 * (CLcr / Median CLcr)

V2 (L) 7.21 0.123 2 14

WGT (V2) 0.0996 0.00626 6 -

V2 (WGT) = 7.21 + 0.0996 * (Weight - Median Weight)

k a (1/h) 1.58 0.184 12 71

Q (L/h) 0.145 0.0143 10 -

V3 (L) 17.8 3.78 21 138

Residual variability 5.78 10 -3 mg/l. CL: Clearance from first compartment; Q: Inter-compartmental Clearance; V2: Distribution Volume of first (central) compartment; V3: Distribution Volume of second

(peripheral) compartment; k a: Rate of absorption. The covariance of CL and V2 within the omegablock was 0.0276, correlation=0.84.

ResultsA total of 20,000 patients were simulated to receive fondaparinux 2.5 mg for 28 days with samplescollected at appropriate time points on Day 7 and Day 28. No specific simulation was run at the1.5 mg dose level: same C max and AUC values obtained at 2.5 mg dose level were normalised to1.5 mg. Empirical 5th, 50th and 95th percentiles were computed for a unit of CL crea every 10 ml/minfrom 20 to 150 ml/min.

In order to provide reasonable estimates of the distribution of simulated Cmax and AUC 0-24 for a givenCL crea value, empirical 5th, 50th and 95th percentiles were computed for a unit of CL crea every 10

m/min. For graphic representations, exponential decay regression curves (y=y 0 + a *e-bx

) wereestimated for each percentile (Fig 3).

7/31/2019 Wc 500027737

27/45

27/45 EMEA 2004

Figure 3 -Individual predicted AUC 0-24 (mg/l) and corresponding 5th , 50 th and 95 th percentiles on

Day 28 at 1.5 mg or 2.5mg once daily with cut-off at Clcrea 30 and 50 l/min in patients

The model shows that increasing the CL crea cut-off to a value above 30ml/min (i.e. 50ml/min)decreases the variability in the population and generates more comparable exposure in RI andnormal patients for a switch between 1.5 and 2.5 mg dose.

7/31/2019 Wc 500027737

28/45

28/45 EMEA 2004

Table 6- Descriptive statistics on predicted C max (mg/l) and AUC 0-24 (mg.h/l) on Day 28 at 1.5 mgand 2.5 mg q.d. for a CL crea value at 20, 30, 40, 50 and 80 ml/min.

Creatinine clearance value

Dose 20 ml/min 30 mL/min 40 ml/min 50 ml/min 80 ml/min

C max (mg/l)

2.5 mg

n 30 56 86 112 235Mean (CV%) 0.760 (30) 0.612 (26) 0.530 (25) 0.456 (23) 0.359 (22)

5th percentile 0.464 0.403 0.325 0.299 0.250

Median 0.714 0.600 0.508 0.459 0.347

95 th percentile 1.149 0.905 0.747 0.630 0.506

1.5 mg

n 30 56 86 112 235

Mean (CV%) 0.456 (30) 0.367 (26) 0.318 (25) 0.274 (23) 0.215 (22)

5th percentile 0.278 0.242 0.195 0.179 0.150

Median 0.428 0.360 0.305 0.275 0.208

95 th percentile 0.689 0.543 0.448 0.378 0.304

AUC 0-24 (mg.h/l)

2.5 mg

n 30 56 86 112 235

Mean (CV%) 14.51 (33) 11.18 (31) 9.35 (27) 7.73 (24) 5.43 (25)

5th percentile 8.62 6.57 5.76 4.85 3.54

Median 13.95 10.88 8.91 7.81 5.22

95 th percentile 21.92 17.96 14.28 10.62 7.89

1.5 mg

n 30 56 86 112 235

Mean (CV%) 8.71 (33) 6.71 (31) 5.61 (27) 4.64 (24) 3.26 (25)

5th percentile 5.17 3.94 3.46 2.91 2.13

Median 8.37 6.53 5.35 4.68 3.13

95 th percentile 13.15 10.77 8.57 6.37 4.73

The MAH has also looked at controlled MB data from the original MAA and from Penthifra Plus inorder to help establish a cut-off value (Table 7).

7/31/2019 Wc 500027737

29/45

29/45 EMEA 2004

Table 7- MB [number (%) of patients] by CL crea and study categories in fondaparinux MOSLLstudies patients actively treated with fondaparinux

Pre-Operative Studies(ACT2545, 63118,

EFC2698)

Post-Operative Studies(DRI2643, EFC2442,

095-002)

ExtendedProphylaxis(EFC4582)

CL crea Fonda

2.5 mg qdEnox

40 mg qdFonda

2.5 mg qdEnox

30 q12hrFonda

2.5 mg qdPlacebo

20ml/min).

The CPMP believes that, in the absence of clinical data to the contrary, it is preferable to opt for adosing regimen that will give similar exposure in RI as in non-impaired patients, especially during

prolonged prophylaxis, as reflected in the adopted SPC. This is relevant since patients with RI in thelower moderate range may not have achieved steady state with the 2.5 mg dose at the end of the

previously approved 9-day treatment period. The fact that patients with renal impairment may be athigher risk for VTE and in special need of prophylaxis does not justify a higher exposure in this population. It is more reasonable that the VTE risk is a reflection of overall morbidity rather thanrelated to renal impairment per se . Moreover, the safety data available for prolonged prohylaxis in RIstem from study EFC4582 and are not extensive.

NEW INDICATION IN THE TREATMENT OF VTE (new pre-filled syringes of 5mg/0.4ml,7.5mg/0.6ml and 10mg/0.8ml)

Rationale for New IndicationVTE disease is the usual terminology for referring to DVT and PE, two clinical expressions of thesame underlying pathophysiology. The mean annual reported incidence of DVT ranges between 48-180 per 100,000 population. Whereas a lower annual incidence of 23/100,000 is reported for symptomatic PE, asymptomatic defect of pulmonary perfusion is reported in up to 80% of patientssuffering from DVT, especially proximal DVT. Conversely, the presence of DVT could bedocumented in up to 93% of PE patients. In addition, the reported rates of VTE recurrences in patientswith only symptoms of DVT associated or not to asymptomatic PE or in patients with symptoms of PEare similar. These data emphasise the intimate relationship between DVT and PE.

DVT and PE are treated with the same therapeutic approach aiming to prevent further thromboticextension with increased risk for post-thrombotic syndrome and pulmonary embolisation that may belife-threatening or lead to restriction of pulmonary function and/or chronic pulmonary hypertension.Current management of patients with acute VTE requires initial treatment with an anticoagulant

7/31/2019 Wc 500027737

30/45

30/45 EMEA 2004

exhibiting rapid onset of action, such as Unfractionated Heparin (UFH) or Low Molecular WeightHeparin (LMWH), followed by chronic anticoagulation with oral anti-vitamin K agents (AVK) for secondary prevention. The approved UFH regimen for this initial treatment of patients with DVTand/or PE requires activated partial thromboplastin time (aPTT) monitoring due to inter-individualdifferences in dose-response relationship. LMWHs, such as bodyweight-adjusted dose enoxaparinsodium represent therapeutic alternatives for DVT patients with superior or equal efficacy, with somesuggestions for decreased bleeding complications and heparin-induced thrombocytopenia (HIT) typeII, which is a serious complication of heparin therapy. This therapeutic alternative, consisting of onceor twice daily sc administration, allows home treatment for suitable patients.

Unlike fondapanirux, which is chemically synthesised, both UFHs and LMWHs are animal-sourced products with potential for contamination and exposure of patients to immunoallergic reactions and toother non-haemorrhagic adverse drug reactions. There is, therefore, a need to improve the initialanticoagulant therapy of patients with VTE.

This application concerns 3 new product strengths (pre-filled syringes of 5mg/0.4ml, 7.5mg/0.6ml and10mg/0.8ml) developed for a new indication in the treatment of DVT and PE. Various healthauthorities were consulted prior to the finalisation of the clinical development plan. The studies followthe main recommendations made in CHMP Note for guidance (NfG) on clinical investigation of medicinal products for the treatment of VTE disease (CHMP/EWP/563/98) and CHMP scientificadvice provided in September 1999.

1. Part II: Chemical, pharmaceutical and biological aspects

CompositionArixtra 5mg/0.4ml, 7.5mg/0.6ml and 10mg/0.8ml contain fondaparinux sodium as the activeingredient in a concentration of 12.5 mg/ml. The medicinal product is presented as a solution for injection in a prefilled syringe with an automatic needle protection system.

Other ingredients include an isotonic solution of sodium chloride and water for injections. Sodiumhydroxide and hydrochloric acid are used as pH adjusters and sterile nitrogen is used as an aid tofiltration, but is not present in the finished product.The product is packaged in a single dose syringe consisting of a glass barrel with a stainless steelneedle and a rubber needle shield and closed with a plunger stopper. To differentiate between thedifferent doses available, the plunger rod and the automatic needle protection system have differentcolours for each strength.

Active substanceThe drug substance is identical to the one used in the already approved strengths.

Other ingredientsThe same excipients are used as in the already approved strengths. All materials are of non-animalorigin and comply with Ph. Eur. requirements.

Product development and finished productArixtra 5mg/0.4ml, 7.5mg/0.6ml and 10mg/0.8ml have been developed using the same qualitativecomposition, manufacturing process and primary packaging material as the already approvedstrengths. Since the proposed indication requires a higher dose compared to the approved ones, ahigher concentration was needed in order to keep the injection volume below 1 ml. For this reason theconcentration of the solution has been increased to 12.5 mg/ml as compared to the already authorised5 mg/ml and the sodium chloride concentration has been changed from 0.84 to 0.76% in order toobtain similar osmolality (284-285 mOsm/kg) for both fondaparinux sodium solutions.

The manufacturing process is the same as the one employed in the already approved strengths. It is astandard process that involves the following steps: compounding, sterile filtration, aseptic filling,terminal sterilisation, inspection and secondary packaging. Filtration and terminal sterilisation have

7/31/2019 Wc 500027737

31/45

31/45 EMEA 2004

been identified as the critical steps of the manufacturing process and have been appropriately validatedtogether with the proposed holding times. From the results presented it can be concluded that themanufacturing process is reproducible and can consistently produce a product that meets the proposedspecifications.

Product specificationThe specifications include tests for the appearance, identification (LC), particulate matter, pH,extractable volume, degradation products (LC), sterility, bacterial endotoxins and assay by validatedanalytical methods. They have been kept identical with the ones for the approved strenghts exceptfrom some minor changes concerning the colour, extractable volume and assay.

Stability of the product Stability data from 3 batches of the 5mg/0.4ml strength, 1 batch of the 7.5mg/0.6ml strength and 3

batches of the 10mg/0.8ml strength stored at 25C/60%RH and 30C/60%RH for 12 months and at40C/75%RH for 6 months have been presented. The samples were tested for appearance, particulatematter, pH, extractable volume, degradation products, free sulfates, sterility, bacterial endotoxins andassay, by validated stability indicating methods. In all cases the batch analysis results met the

proposed specification.Data from photostability and temperature cycling studies conducted according to ICH guidelines havealso been provided. All results remained within the acceptance criteria. The stability data provided indicate that the medicinal product is stable for the proposed shelf lifewhen stored under the conditions specified in the SPC.

Discussion on chemical, pharmaceutical and biological aspects.The quality of Arixtra 5mg/0.4ml, 7.5mg/0.6ml and 10mg/0.8ml is adequately established. In general,satisfactory chemical and pharmaceutical documentation has been submitted for the marketingauthorization. There are no major deviations from EU and ICH requirements.

The proposed new strengths are similar to the already approved ones. They have the same qualitativecomposition and only differ slightly in quantitative composition. The same manufacturing process and

primary packaging material are used as in the approved strengths and stability tests indicate that the product under ICH guidelines conditions is chemically stable for the proposed shelf life.

2. Non-clinical aspects

2.1 Toxicology

The mean maximum human exposure is about three-fold higher with the new proposed dosageregimen than with the hitherto approved posology. No new preclinical information has been madeavailable since the product was originally approved. Further to a request from CHMP, the Applicanthas submitted a non clinical overview including animal/human AUC-exposure ratios with the new

proposed posology, together with a justification as to why the available preclinical information issufficient to support the safety of the product at the higher proposed doses. Thus, the exposure ratios

between the highest dosed animals of the pivotal repeat dose toxicity studies and patients with normalrenal function ( 20 gxh/ml) have decreased from 5 to 1.75 in rat, 13 to 4.25 in rabbit, and from 16 to5.3 in monkey. In renally impaired patients the exposure ratios will be even lower (1, 2.4 and 3 with

the applicants dose-adjustment proposal given 35 gxh/ml).

As explained in the Toxico-Pharmacological Critical Assessment of the original application, acombination of safety, pharmacokinetic and ethical considerations justified why studies with doseshigher than those used cannot be performed; consequently, no additional toxicological studies withhigher doses have been performed. The Applicant refers to this justification and argues that theanimal/human exposure ratios obtained for the recommended therapeutic treatment dose of 7.5 mg areadequate and sufficient to support the safety of Arixtra for the intended indication, and concludes thatno changes in the preclinical safety sections of the SmPC are necessary.

7/31/2019 Wc 500027737

32/45

32/45 EMEA 2004

Discussion on the non-clinical aspects

The investigated margins of toxicity are generally low. The question arises if more can be done tocover the new higher exposure scenario. In the earlier submitted repeat dose/reproduction toxicitystudies (max dose 10 mg/kg/day in all species), the maximum tolerated dose (MTD) appears not tohave been reached in the rat and rabbit (mostly minor injection site haemorrhage reported) in contrastto the more exposed monkeys (up to 3-month iv treatment), where a relatively more severe toxicity

profile of primary bleeding and secondary associated events was reported. Hence, the full feasibletoxicity profile of fondaparinux appears to have been explored in a species considered relevantregarding the pharmacological effects of heparins on the coagulation system. There were no apparentindications of other toxic events in the monkey, considered unrelated to effects on the coagulationsystem, up to 3-5 fold the clinical exposure level. The monkey data indicate that, upon short-termtreatment, unacceptable bleeding would arise before any other potential harmful reaction occurs.Given the MTD studies available in monkeys, the CHMP concurs that it is not necessary to collectfurther data in the rodent species. Nonetheless, in contrast to the applicants view, the CHMPconsiders that the toxicological profile of fondaparinux is not fully characterised for this higher doseindication. Hence, the fact that the repeated dose and reproduction toxicity studies are insufficient dueto the limited exposure has been appropriately reflected in section 5.3 of the SPC.

3. Clinical aspects

3.1 Introduction

The efficacy and safety of Arixtra 5mg/0.4 ml, 7.5 mg/0.6 ml, and 10 mg/0.8 ml solution for injectionin the treatment of Venous Thromboembolic Events was investigated in 3 multicentre studiesconducted in Europe, North America and Australia, and additional pharmacokinetic (PK) and

pharmacodynamic (PD) data were collected. The phase II REMBRANDT study (DRI2440) was adose-response study conducted in patients with a bodyweight of 50-100 kg and tested 5 mg, 7.5 mgand 10 mg once daily sc fondaparinux versus q12h sc dalteparin in the initial treatment of symptomatic proximal DVT. Two pivotal phase III non-inferiority studies using the proposed weightadjusted dosing scheme (5mg 100kg) were performed: one double-

blind enoxaparin-controlled study in the curative treatment of DVT (MATISSE-DVT- studyEFC2441) and one open-label UFH-controlled study in the curative treatment of PE (MATISSE-PE-study 63121). All studies were performed in compliance with the GCP regulations effective in Europe,

North America and Australia. In addition, clinical protocols, amendments to the protocols, patientinformed consents were reviewed and approved by independent Ethics Committee or InstitutionalReview Board.

Overall, the development program randomised 4,874 patients, as follows.

7/31/2019 Wc 500027737

33/45

33/45 EMEA 2004

Randomized PatientsN = 4874

Fondaparinux sodium 5 mg

N = 104

Fondaparinux sodium 10 mg

N = 121

Fondaparinux sodium 7.5 mg

N = 111

Fondaparinux sodiumproposed regimen

N = 2312

Randomized Patients in fondaparinux sodium group

N = 2537 Randomized Patients in comparator groups

N = 2337

DRI 2440 N = 336 63123N = 1103 EFC 2441N = 1098 DRI 2440

( dalteparinsodium) N =120

63123

(UFH) N =1110

EFC 2441 ( enoxaparin sodium)

N =1107

Randomised PatientsN = 4874

Fondaparinux sodium 5 mg

N = 104

Fondaparinux sodium 10 mg

N = 121

Fondaparinux sodium 7.5 mg

N = 111

Fondaparinux sodiumproposed regimen

N = 2312

Randomised Patients in fondaparinux sodium group

N = 2537 Randomised Patients in comparator groups

N = 2337

DRI 2440 N = 336 63123N = 1103 EFC 2441N = 1098 DRI 2440 N = 336 63123N = 1103 EFC 2441N = 1098 DRI 2440

( dalteparinsodium) N =120

63123

(UFH) N =1110

EFC 2441 ( enoxaparin sodium)

N =1107

DRI 2440 ( dalteparinsodium)

N =120

63123

(UFH) N =1110

EFC 2441 ( enoxaparin sodium)

N =1107

3.2 Pharmacokinetics

BioequivalenceThe clinical formulation employed in the phase II and phase III Matisse-DVT study was identical tothat used for the clinical trials presented in the original prophylaxis dossier. In the formulation used for the other phase III study (Matisse-PE), the fondaparinux sodium concentration was increased to 12.5mg/ml in anticipation of the market formulation. The bioequivalence of the new formulation and thereference formulation has been shown with respect to both rate and extent of absorption.

Pharmacokinetics in target populationThe pharmacokinetics were evaluated on Days 1 and 5. The results show a tendency towards a non-

proportional dose-concentration relationship (AUC is increased by 50% when doubling the dose)indicating that a non-linear process is occurring. It has been previously shown in vitro that protein

binding is saturated above 2 mg/l. Since the effect and bleeding events of fondaparinux are related todrug bound to antithrombin, saturated binding could hypothetically result in approaching the maximaleffect. Further to a request from CHMP, the Applicant has provided data showing that, within theconcentration range achieved with 5-10 mg fondaparinux once daily (up to 3 mg/l), AT III binding isnot saturated. Increasing the dose further may result in a saturation of AT III binding.

A comparison of steady state plasma concentrations observed in patients with DVT at a regimen of 10 mg fondaparinux (study DRI2440) with data obtained in healthy volunteer studies (previousdossier) indicates a close similarity in the PK of fondaparinux in the 2 populations, suggesting that thePK observations in previous studies in healthy volunteers hold true for the patient population under discussion.

In addition, a detailed NONMEM population PK analysis was conducted pooling all PK data obtainedin studies DRI2440, EFC2441 and 63121 investigating numerous possible covariate relationships(gender, age, body weight, renal function, etc) and their effect on the kinetics of fondaparinux. Renalfunction was identified as the most important covariate affecting clearance, and the simulationsdemonstrated that with dosage adjusted to renal function, variability in exposure can be reduced.Hence, for fondaparinux, dosage primarily according to renal function would be the optimal posology.However, given the data available, the CHMP did not find that a change from the weight-based dosageused in clinical trials to renal function-based dosage would be possible in the present application.

Having identified renal function as the most important variable affecting exposure to fondaparinux,the CHMP raised several questions to better understand the kinetic and dynamic findings in patientswith moderate or severe renal impairment. As seen from the applicants response, the numbers of

7/31/2019 Wc 500027737

34/45

34/45 EMEA 2004

subjects with CLcr 30-40 and 40-50 ml/min are fairly large (129 and 189, respectively). Althoughthere is a clear trend to increased incidence of VTE and bleedings in severe renal impairment (CLcr 20-30 ml/min) see below, the data for patients with moderate renal impairment do not raise newconcerns and do not reveal any clear differences between patients with CLcr 30-40 and 40-50 ml/min.It should be noted, however, that the majority of patients with moderate renal impairment had a bodyweight of 50-100 kg and that there is truly limited information in patients with low body weight andmoderate renal impairment, and no information in patients with high body weight and moderate renalimpairment.

Patients with Severe renal impairment (CL cr

7/31/2019 Wc 500027737

35/45

35/45 EMEA 2004

presence of concomitant VKA therapy. The changes in aPTT seem dose-dependent but moderate, of similar amplitude to those observed with dalteparin and difficult to interpret in the presence of concomitant VKA treatment. Thus, monitoring of the pharmacodynamic activity of fondaparinuxsodium by aPTT does not seem appropriate.

Fondaparinux sodium has been shown in healthy subjects not to interact with the pharmacokineticsand pharmacodynamics of warfarin. During the VTE treatment studies, the pharmacodynamic activityof VKA (started within 72 hours of randomisation) was assessed by INR. No differences between thetreatment groups were recorded in either the mean INR values achieved at the end of the study drugadministration period or in the time to reach an INR 2.0. Somewhat more fondaparinux treated

patients had an INR >3 at the time of study drug discontinuation as compared with the active controls.Unfortunately, the mean doses of VKA at discontinuation of study drug in the different treatmentgroups, requested by CHMP, were not recorded during the studies. The Applicant has subsequentlyclarified that fondaparinux increases the prothrombin time to a slight extent, an effect judged to be of minor clinical relevance.

Age and renal function play a major role on fondaparinux elimination and there is an increased risk of dangerously high levels in these fragile populations. Further to a request from CHMP, the Applicanthas addressed the question of whether indirectly monitoring fondaparinux concentrations bymeasuring plasma factor Xa levels could be considered in fragile patients. The Applicant refers to thedifficulties in defining a therapeutic range and concludes that it is not useful. However, it may inselected cases be regarded as desirable by the treating physician to have the possibility to check the

plasma levels achieved in fragile patients, especially when treatment is prolonged. The situation can be compared to that of LMWH, where monitoring of plasma levels is seldom part of clinical routine but can be easily determ